Risk of suicidal behaviors following sport-related and non-sport-related concussion: a systematic review and meta-analysis

Abstract

Objective

This systematic review and meta-analysis aimed to quantify the strength of association between a history of concussion and the spectrum of suicidal behaviors, including Suicidal Ideation, Made a Suicide Plan, Attempted Suicide, and Death by Suicide. It also sought to provide the first direct comparison of suicide risk associated with sport-related concussion (SRC) versus non-sport-related concussion (non-SRC) and to examine the moderating role of sex.

Methods

Following the PRISMA guidelines, this study was prospectively registered with PROSPERO (CRD420251074749). A systematic search of PubMed, Embase, Cochrane Library, Web of Science, and EBSCO databases was conducted through August 2025. Observational studies reporting on the association between a history of concussion and at least one suicidal behavior outcome were included. A random-effects model was used to calculate pooled odds ratios (ORs) with 95% confidence intervals (CIs). Subgroup analyses were performed based on the context of injury (SRC vs. non-SRC) and sex.

Results

A total of 18 studies, comprising 13,483,226 participants (of whom 650,465 had a history of concussion), were included. The meta-analysis revealed that a history of concussion was significantly associated with an increased risk for all suicidal behavior outcomes: Suicidal Ideation (OR = 1.35, 95% CI: 1.20–1.52), Made a Suicide Plan (OR = 1.35, 95% CI: 1.20–1.51), Attempted Suicide (OR = 1.66, 95% CI: 1.43–1.94), and Death by Suicide (OR = 1.90, 95% CI: 1.59–2.27). The strength of this association increased in a graded manner corresponding to the severity of the suicidal behavior. Subgroup analysis stratified by sex revealed that for Attempted Suicide, the risk was significantly higher in males with a history of concussion (OR = 2.03) compared to females (OR = 1.41). Furthermore, the degree of risk increase for suicidal behaviors brought by sport-related and non-sport-related concussion was comparable and both were statistically significant (p < 0.05).

Conclusion

Concussion, whether sport-related or non-sport-related, is a risk factor for suicidal behaviors, including Suicidal Ideation, Made a Suicide Plan, and Attempted Suicide, with comparable levels of risk for both contexts. Given the elevated risk for Attempted Suicide observed among males following a concussion, the integration of suicide risk screening and sex-specific intervention strategies into the standard of care for all patients with concussion is warranted.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12888-025-07509-5.

Introduction

Concussion is a temporary brain dysfunction caused by trauma, which mainly originates from two major contexts, sport-related activities and non-sport-related events such as falls, assaults, or motor vehicle accidents. It is an important public health problem, with millions of people affected each year; among them, sport-related concussions (SRCs) are particularly common in adolescents and young athletes, while non-sport-related concussions affect a broader population, including young children and the elderly [1, 2]. Although concussion is generally considered a self-limiting injury, a considerable portion of patients experience physiological, cognitive, and emotional symptoms that persist for months or even years [3–5]. These chronic aftereffects can lead to long-term impairment of learning ability, functional capacity, and overall quality of life [6, 7]. However, a key limitation in clinical practice remains the lack of evidence-based therapies that can effectively prevent or mitigate the long-term neuropsychiatric sequelae of concussion, leading clinicians to primarily rely on supportive care strategies [8, 9].

Among the many sequelae of concussion, its astonishing association with suicidal behavior is receiving increasing attention. A large number of studies have shown that individuals with a history of concussion have a risk of experiencing Suicidal Ideation, Attempted Suicide, and even Death by Suicide that is about twice that of the population without a history of concussion [10], a phenomenon that is reflected in military personnel, adolescents, and athletes [2, 11–14]. Furthermore, concussion is also associated with hopelessness and increased capability for suicide. These findings emphasize the importance of concussion prevention, early intervention, and mental health support for people with a history of concussion, especially adolescents and athletes. The mechanism behind this association is considered to be multifactorial; brain trauma can directly cause neurobiological damage, leading to elevated levels of inflammatory cytokines closely related to suicidal behavior [15, 16]; at the same time, the persistent symptoms and functional impairment brought by concussion may also induce or exacerbate psychological distress such as depression, anxiety, and feelings of hopelessness, which are themselves strong independent risk factors for suicide. Early identification and treatment of depression and anxiety in patients with concussion are crucial for optimizing recovery and preventing suicide [17, 18].

Although the association between concussion and suicide risk has received widespread attention, several key knowledge gaps exist in the current body of evidence, limiting our refined understanding of the risk and the specificity of clinical interventions. Previous comprehensive meta-analyses, such as the one conducted by Fralick et al. in 2019, failed to conduct separate quantitative assessments for different levels of suicidal behavior like Suicidal Ideation, plans, and attempts, due to extremely high heterogeneity among the included studies (I²=96%) [1]. Secondly, a long-standing key question in clinical practice and research is: are sport-related concussion (SRC) and non-sport-related concussion equivalent in terms of the risk of causing neuropsychiatric sequelae? There is currently a lack of direct, comprehensive meta-analytic evidence to answer this question, leading to a potential bias in clinical settings of viewing sport-related concussion as a “milder” injury. Finally, although sex differences are a recognized important factor in suicide behavior research, in the field of suicide risk after concussion, previous studies have rarely provided systematic quantitative evidence, especially lacking in-depth analysis of the moderating effect of sex on different suicide outcomes. To fill the above gaps, this meta-analysis aims to make unique contributions to the existing evidence base in the following three aspects: first, to systematically and stratigraphically evaluate for the first time the association strength between a history of concussion and four different outcomes in the suicidal behavior spectrum (Suicidal Ideation, Made a Suicide Plan, Attempted Suicide, Death by Suicide); second, to directly compare for the first time the suicide risk brought by sport-related and non-sport-related concussions to test whether injury context is a key risk moderating factor; third, to conduct quantitative subgroup analysis on the moderating role of sex in different suicidal behaviors. By addressing these specific knowledge gaps, this study aims to provide higher-quality evidence-based medicine evidence for clinical risk stratification, early screening, and the development of more targeted prevention strategies.

Methods

This research analysis was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines [19]. The study protocol was pre-registered on the International Prospective Register of Systematic Reviews (PROSPERO), with registration number CRD420251074749.

Data sources and searches

On August 1, 2025, this study conducted a comprehensive, systematic search of five major electronic databases, including PubMed, Embase, Cochrane Library, Web of Science, and EBSCO. The search scope covered all literature from each database’s inception date to the search date. The search strategy was built around three core concepts to ensure precise retrieval of relevant studies: exposure (concussion), outcome (suicide), and risk/association. The search terms for exposure included a combination of Medical Subject Headings (MeSH) and keywords, such as (“Brain Concussion“[Mesh] OR “Brain Injuries“[Mesh] OR Concussion* OR “brain injur*” OR “brain traum*” OR “cortical contusion*” OR “Commotio Cerebri*”). The search terms for outcome included (“Suicide“[Mesh] OR Suicidal* OR Suicide* OR “self kill*”). To improve the retrieval specificity of epidemiological studies, a third concept of risk was added, using terms like (“Risk“[Mesh] OR Risk*). Within each concept, search terms were connected using the “OR” Boolean operator; between the three core concepts, the “AND” operator was used. The complete, unabridged search strategies for all five databases are detailed in the supplementary materials (Supplementary Table 1). Additionally, we also manually screened the reference lists of all included studies and relevant reviews to identify any other eligible literature.

Eligibility criteria

Studies had to meet the following criteria to be included: (1) Study design: The study was an observational study, including retrospective or prospective cohort studies and cross-sectional studies. (2) Study population: The study included an exposed group with a clearly diagnosed history of concussion and a control group without a history of concussion, with no restrictions on population type (e.g., students, military personnel, general population), age, or sex. (3) Exposure factor: The exposure of interest was concussion, regardless of etiology (i.e., sport-related or non-sport-related). (4) Outcome measures: The primary outcomes were suicidal behaviors, specifically defined as Suicidal Ideation, Made a Suicide Plan, Attempted Suicide, or Death by Suicide. (5) Data requirements: The study must report an effect size quantifying the association between concussion and suicidal behavior, and its corresponding 95% confidence intervals (CIs), or provide raw data from which these estimates could be calculated. Conference abstracts, review articles, study protocols, or case reports would be excluded. If multiple publications reported on the same cohort, the study with the longest follow-up or the largest sample size was included. Studies that did not report relevant outcome measures or from which the necessary effect estimates could not be extracted were also excluded. At the same time, only studies that explicitly identified concussion were included, excluding studies on moderate-to-severe traumatic brain injury, penetrating injuries, or mixed-severity samples where data specific to concussion could not be distinguished.

Study selection

Two researchers (MYL and ELH) independently conducted literature screening according to the pre-set inclusion and exclusion criteria. First, duplicates and obviously irrelevant studies were excluded by reading titles and abstracts. Subsequently, full-text assessment was performed on potentially eligible literature to determine final inclusion. Any disagreements were resolved by reaching a consensus through discussion with a third researcher (KYL).

Risk of bias

This study used the criteria for cross-sectional studies recommended by the Agency for Healthcare Research and Quality (AHRQ) [20] to assess the methodological quality of the included studies. Two researchers (MYL and ELH) independently evaluated the quality of each study and cross-checked the results. Disagreements were resolved through discussion, and a third researcher was consulted if necessary. The AHRQ tool includes 11 items, with a total score ranging from 0 to 11 points. According to the score, studies were classified as low quality (0–3 points), moderate quality (4–7 points), or high quality (≥ 8 points). For the included cohort studies, their methodological quality was evaluated using the Newcastle-Ottawa Scale (NOS) [21]. This scale assesses three key domains: selection and exposure assessment of study subjects (maximum 4 points), comparability between study groups (maximum 2 points), and outcome assessment and adequacy of follow-up (maximum 3 points). Based on the total score, studies were classified as low quality (0–3 points), moderate quality (4–6 points), or high quality (7–9 points). Detailed assessment results are shown in the supplementary materials (Tables 1 and 2).

Table 1.

Characteristics of included studies

Author	Country	Study Period	Design	Study Sample	Concussion Sample Size	Non-concussion Sample Size	Ascertainment Method and Injury Type	Relevant Adjustment Factors
Fogarty et al., 2024	US	January to March 2021	cross-sectional study	High school students	547	3889	YRBSS, Sports-Related Concussion	Age, race/ethnicity, sexual minority orientation, and bullying victimization experience
Wang et al. 2024	China	From October 26 to November 18, 2021	cross-sectional study	Chinese university students	6267	37,602	SBQ-R, Concussion	Including symptoms of depression and anxiety, history of bullying victimization, and childhood maltreatment.
Kay et al., 2023	US	2017 and 2019	cross-sectional survey.	United States secondary school students	3431	20,014	YRBS, sport- or recreation-related concussion	Age, race or ethnicity, and alcohol use.
Bailey et al., 2023	US	2019	cross-sectional study	high school students	1897	10,725	YRBS, Concussion	None
Ledoux et al., 2022	Canada	between April 1, 2010, and March 31, 2020	retrospective cohort study	children and youths aged 5 to 18 years	152,321	296,482	ICD-10-CA, Concussion	Residential neighborhood income quintile, child abuse or neglect, migraine history, organic mental disorders or developmental
Eagle et al. 2022	US	From 2017 to 2019	cross-sectional study	youths aged up to 18 years	3874	24,568	YRBS, Sports-Related Concussion	depression history and age
Iverson and Karr, 2022	US	2019	cross-sectional study	high school students	1442	9820	YRBS, Concussion	Age, race or ethnicity
Ziminski et al., 2022	US	2017 and 2019	cross-sectional study	adolescents	1965	12,531	YRBS, Sports-Related Concussion	Gender, age, race/ethnicity, risk, and protective factors,
Mantey et al., 2021	US	2017	cross-sectional study	11–14 years old students	3461	15,874	YRBS, Concussion	Sex, age, race/ethnicity,
Miller et al., 2021	US	2017	cross-sectional study	US High School Students	1057	6828	YRBS, Sports-Related Concussion	Sex, grade level, and race/ethnicity
Campbell-Sills et al. 2020	US	30month	prospective cohort study	US Army Soldiers	4808	2869	PPDS, PCS-8, concussion and mild TBl	Age, sex, high school or less education, marital/relationship status of “previously married, not currently engaged/in a serious relationship”, having deployed within the first year of Army service, perceived unit cohesion, lifetime mental disorder
Knell et al. 2020	US	2017	cross-sectional study	American High School Athletes	628	14,604	YRBS, Concussion	None
Wangnoo et al., 2020	US	2013	cross-sectional study	12th grade high school students	531	1614	YRBS, Concussion	None
Mantey et al., 2020	US	2017	cross-sectional study	high school students	2003	11,350	YRBS, Sports-Related Concussion	Race/ethnicity, grade level, sexual orientation, and history of bullying victimization
Yang et al., 2019	US	2017	cross-sectional study	Physically Active High School Students	664	2763	YRBS, Concussion	None
Fralick et al., 2016	Canada	1992–2012	retrospective cohort study	General population	235,110	2,397,192	ICD-9 and ICD-10 codes, Concussion	None
Fazel et al., 2014	Sweden	1969–2009	retrospective cohort study	General population	218,300	2,163,190	ICD codes (unspecified edition), concussion	Age, sex, income, marital status, immigration status
Brenner et al., 2011	US	2001–2006	retrospective cohort study	Military VHA	12,159	7,800,846	ICD-9 codes, concussion	None

Table 2.

Subgroup analysis of suicide risk after concussion

Suicidal Risk	Subgroups		Included studies	OR(95%CI)	I2(%)	P-values
Suicidal Ideation	Injury Context	Non-Sport-Related Concussions	6	1.38(1.12–1.71)	89.70%	0.003
		Sports-Related Concussion	5	1.34(1.26–1.43)	0.00%	0.000
	Sex	Male	6	1.25(1.10–1.43)	25.30%	0.001
		Female	6	1.32(1.21–1.45)	0.00%	0.000
Made a Suicide Plan	Injury Context	Non-Sport-Related Concussions	6	1.34(1.08–1.65)	87.3%	0.007
		Sports-Related Concussion	5	1.34(1.25–1.44)	0.00%	0.000
	Sex	Male	5	1.17(0.99–1.38)	39.40%	0.001
		Female	5	1.31(1.14–1.51)	31.80%	0.000
Attempted Suicide	Injury Context	Non-Sport-Related Concussions	7	1.61(1.20–2.15)	91.00%	0.001
		Sports-Related Concussion	6	1.70(1.45–1.99)	80.60%	0.000
	Sex	Male	5	2.03(1.47–2.81)	76.90%	0.000
		Female	5	1.41(1.19–1.67)	47.90%	0.000

Statistical analysis

All statistical analyses were performed using Stata 14.0 software (Stata Corp, College Station, TX). To assess the association between a history of concussion and the risk of suicidal behavior, adjusted odds ratios (aOR) and their 95% confidence intervals (CI) were preferentially extracted and pooled from each study. When a study reported multiple adjustment models, the model that included the most comprehensive clinically relevant covariates (e.g., demographic factors, mental health history, and victimization experiences) was selected to obtain the most robust estimate. If a study only provided raw data, unadjusted ORs were calculated based on exposure and outcome events [22]. Given the clinical and methodological differences among the studies, the primary analyses all used a random-effects model to pool effect sizes [23, 24]. Heterogeneity was measured by the Cochran’s Q test (P < 0.05) and the I² statistic, where I² < 25%, 25–50%, 50–75%, >75% represented no, low, moderate, and high heterogeneity, respectively [25]. To assess the robustness of the pooled effect estimate, we conducted a “leave-one-out” sensitivity analysis [26]. Sensitivity analysis was performed by deleting studies one-by-one to assess the impact of all included studies on the overall risk of suicidal behavior. The assessment of publication bias was performed through visual inspection of the funnel plot combined with Egger’s regression test [27]. When potential publication bias was detected, the “trim-and-fill method” was used to adjust for the potential publication bias [28, 29]. In addition, to explore potential sources of heterogeneity, we conducted subgroup analyses based on pre-specified variables, namely concussion type (sport-related vs. non-sport-related) and sex (biological sex). Among them: SRC is defined as a traumatic brain injury caused by a direct blow to the head, neck, or body, resulting in force transmitted to the brain, in sports and exercise-related activities [30]; non-SRC is defined as a concussion occurring outside of sports activities caused by mechanisms such as falls, slips or trips, being struck by an object, motor vehicle accidents, and assaults [31]; Sex (physiological sex) refers to the biological and physiological characteristics (chromosomes, gonads, reproduction, anatomical structures) that define humans as female, male, or intersex [32].

Results

Literature search

This study conducted a systematic search of studies published up to August 1, 2025, in five databases (PubMed, Embase, Cochrane Library, Web of Science, and EBSCO). The search strategy combined search terms for concussion (e.g., “Brain Concussion“[Mesh], brain injur*) and suicide (e.g., “Suicide“[Mesh], Suicidal*). This process initially identified 4,797 records. After removing 1,279 duplicate records, we screened the titles and abstracts of the remaining 3,518 records, from which 388 articles were selected for full-text review. After a comprehensive assessment, a final of 18 studies [12, 13, 33–48] met the inclusion criteria and were included in this meta-analysis. The detailed literature screening process is shown in Fig. 1.

Fig. 1.

Studies screening process

Study characteristics

A total of 18 studies were included in this analysis (Table 1), which included 13 cross-sectional studies [12, 13, 33–37, 39–42, 44, 45], 4 retrospective cohort studies [38, 46–48], and 1 prospective cohort study [43]. Most studies were conducted in the United States, with the remaining studies conducted in Canada [38, 46], Sweden [47], and China [33], respectively. The total sample size of all studies was 13,483,226 participants, of whom 650,465 had a history of concussion and 12,832,761 did not. The geographic distribution of the studies shows that most studies were conducted in the United States, with two others from Canada, and one from Sweden. The study populations were broad in coverage, specifically including adolescents and student groups, military personnel, and the general population. The sample sizes of the studies varied significantly, with the sample size of the concussion group ranging from 531 to 235,110, and the control group without concussion ranging from 1,614 to 7,800,846. The diagnosis of concussion was mainly based on two types of methods. Several cross-sectional studies used validated self-report questionnaires [33, 43], such as the Youth Risk Behavior Surveillance System (YRBSS) [12, 13, 34–37, 39–42, 44, 45]. In contrast, several cohort studies relied on standardized diagnostic codes, such as the International Classification of Diseases (ICD-9 or ICD-10) for case confirmation [38, 46–48]. Most studies performed statistical adjustment for potential confounding factors. Common adjustment variables included demographic information, such as age, sex, race/ethnicity [34, 36, 39]. Some studies also further controlled for mental health history (e.g., depression, anxiety symptoms) and victimization experiences (e.g., school bullying) [13, 33, 40]. For example, Ledoux et al. (2022) adjusted for factors such as history of child abuse and history of migraine in their study [38], while Campbell-Sills et al. (2020) comprehensively adjusted for multiple variables such as marital status, military service history, and unit cohesion in their prospective study [43]. Nevertheless, six studies did not explicitly report or did not perform confounding factor adjustment [12, 35, 44–46, 48].

Literature quality evaluation

The Newcastle-Ottawa Scale (NOS) was used to assess the methodological quality of cohort studies, and the Agency for Healthcare Research and Quality (AHRQ) tool was used for cross-sectional studies. The assessment results showed that all five cohort studies were of moderate to high quality, with NOS scores ranging from 7 to 8. For cross-sectional studies, the AHRQ tool scores ranged from 5 to 9. A total of eighteen studies, including all five cohort studies and thirteen cross-sectional studies, had an average quality score of 6.9, indicating that the evidence base of this meta-analysis has a moderate to high overall methodological quality. Detailed assessment results are shown in the supplementary materials (Tables 1 and 2).

Concussion and suicidal ideation

A total of 11 studies evaluated the association between a history of concussion and Suicidal Ideation. The analysis showed significant heterogeneity among these studies (I² = 84.%, p < 0.001), so a random-effects model was used for the analysis. The pooled result showed that individuals with a history of concussion had significantly higher odds of Suicidal Ideation (OR = 1.35, 95% CI: 1.20–1.52, p < 0.001) (Fig. 2).

Fig. 2.

Meta-analysis of the risk of suicidal ideation after concussion

Concussion and made a suicide plan

The meta-analysis of 11 studies on the association between concussion and Made a Suicide Plan showed significant heterogeneity among the studies (I² = 78.4%, p < 0.001), therefore a random-effects model was used. The pooled result indicated that a history of concussion was associated with a significantly increased odds of Made a Suicide Plan (OR = 1.35, 95% CI: 1.20–1.51, p < 0.001) (Fig. 3).

Fig. 3.

Meta-analysis of the risk of made a suicide plan after concussion

Concussion and attempted suicide

The meta-analysis of 13 studies on Attempted Suicide showed significant heterogeneity among the studies (I² = 87.1%, p < 0.001), thus a random-effects model was required. The pooled result showed that a history of concussion was associated with a significantly elevated odds of Attempted Suicide (OR = 1.66, 95% CI: 1.43–1.94, p < 0.001) (Fig. 4).

Fig. 4.

Meta-analysis of the risk of suicide attempts after concussion

Concussion and death by suicide

Four studies examined Death by Suicide after concussion, with moderate heterogeneity between studies (I² = 55.0%, p = 0.083). The pooled analysis indicated that the risk increased by 90% (OR = 1.90, 95% CI: 1.59–2.27, p < 0.001) (Fig. 5).

Fig. 5.

Meta-analysis of the risk of death by suicide after concussion

Subgroup analysis

Suicidal ideation

A history of concussion was associated with a significant increase in the odds of Suicidal Ideation for both males (OR = 1.25, 95% CI: 1.10–1.43) and females (OR = 1.32, 95% CI: 1.21–1.45). Heterogeneity in the male subgroup was (I² = 25.30%), while the female subgroup had no heterogeneity (I² = 0.00%). In terms of injury context, both non-sport-related concussion (OR = 1.38, 95% CI: 1.12–1.71) and sport-related concussion (OR = 1.34, 95% CI: 1.26–1.43) were associated with higher odds. The sport-related concussion subgroup showed no heterogeneity (I² = 0.00%), whereas the non-sport-related concussion subgroup showed significant heterogeneity (I² = 89.70%) (Table 2).

Made a suicide plan

Females with a history of concussion had a significantly higher risk of Made a Suicide Plan (OR = 1.31, 95% CI: 1.14–1.51), with low-to-moderate heterogeneity (I² = 31.80%). In contrast, this association did not reach statistical significance in males (OR = 1.17, 95% CI: 0.99–1.38). In the analysis by injury context, both non-sport-related concussion (OR = 1.34, 95% CI: 1.08–1.65) and sport-related concussion (OR = 1.34, 95% CI: 1.25–1.44) were associated with a significant increase in risk. However, there was a significant difference in heterogeneity between the two, with no heterogeneity in the sport-related concussion subgroup (I² = 0.00%), while the non-sport-related concussion subgroup showed high heterogeneity (I² = 87.3%) (Table 2).

Attempted suicide

The risk of Attempted Suicide was significantly elevated for both males (OR = 2.03, 95% CI: 1.47–2.81) and females (OR = 1.41, 95% CI: 1.19–1.67), with the risk increase being particularly prominent in males. After stratification by injury context, both non-sport-related concussion (OR = 1.61, 95% CI: 1.20–2.15) and sport-related concussion (OR = 1.70, 95% CI: 1.45–1.99) were associated with a significantly elevated risk of Attempted Suicide. Both subgroups showed high heterogeneity (I² = 91.00% and I² = 80.60%, respectively) (Table 2).

Publication bias

The funnel plot and Egger’s test (p > 0.05) results showed that no significant publication bias was found in the outcome analyses for Made a Suicide Plan, Attempted Suicide, and Death by Suicide (Fig. 6C, D, E). However, in the analysis of Suicidal Ideation, the funnel plot appeared asymmetrical (Fig. 6A), and the Egger’s test result was significant (p = 0.029), suggesting possible publication bias. After correction with the “trim-and-fill method” (Fig. 6B), the pooled effect size was still significant (OR = 1.14, 95% CI: 1.013–1.285), confirming that the publication bias did not produce a substantial change to the study results.

Fig. 6.

Funnel plots of included studies

Sensitivity analysis

To assess the robustness of the results, a sensitivity analysis of excluding studies one-by-one was conducted. As shown in Supplementary Figs. 1–4, after sequentially excluding any single study, the pooled effect sizes and their 95% confidence intervals for the four outcome measures did not change substantially, and the statistical significance of the results remained unchanged.

Discussion

Principal findings

This comprehensive meta-analysis included 18 studies and provided evidence for the association between a history of concussion and an increased risk of suicidal behavior, from Suicidal Ideation to Death by Suicide. The main finding of this study is that for individuals with a history of concussion, their odds of Suicidal Ideation are 35% higher (OR = 1.35, 95% CI: 1.20–1.52), the odds of Made a Suicide Plan are 35% higher (OR = 1.35, 95% CI: 1.20–1.51), the odds of Attempted Suicide are 66% higher (OR = 1.66, 95% CI: 1.43–1.94), and the odds of Death by Suicide are 90.0% higher (OR = 1.90, 95% CI: 1.59–2.27). These results present a clear risk gradient, that is, the strength of the association increases as the severity of the suicidal behavior increases, reaching a peak in the two most severe outcomes, Attempted Suicide and Death by Suicide. It was learned through subgroup analysis that both sport-related concussion and non-sport-related concussion have significant risk in all outcomes. The association between sport-related concussion (SRC) and Suicidal Ideation and plans showed high consistency across studies (I² = 0.00%), while the association for non-sport-related concussion had high heterogeneity (I² >87%). Additionally, the sex subgroup analysis clearly pointed out that males with a history of concussion (OR = 2.03) have a significantly higher risk of Attempted Suicide compared to females (OR = 1.41).

Comparison with previous studies and potential mechanisms

This study found that for severe suicidal outcomes after concussion like Attempted Suicide (OR = 1.66) and Death by Suicide (OR = 1.90), this is basically consistent with the research results of Fralick et al. [2], and enhances the reasonableness of this conclusion by integrating updated evidence. At the same time, this study’s subgroup analysis showed that both sport-related concussion and non-sport-related concussion are associated with a significant increase in suicide risk, and the risk levels are comparable, covering Suicidal Ideation (sport-related OR = 1.34 vs. non-sport-related OR = 1.38), Made a Suicide Plan (both OR = 1.34), and Attempted Suicide (sport-related OR = 1.70 vs. non-sport-related OR = 1.61). This indicates that concussion caused under any circumstances is an important factor that leads to the risk of suicidal behaviors [33, 35].

Furthermore, this study quantified the moderating role of sex in suicide risk, especially in Attempted Suicide, where the risk for males after concussion is almost twice that of females. This finding suggests a specific interaction between males and the sequelae of concussion. This may be mediated by multiple factors, including potential neurobiological differences in the male response to injury, the existence of sex dimorphism in the neuroinflammatory response after concussion [49, 50]; differences in psychosocial factors, as the study by Miyashita et al. (2016) clearly pointed out that male adolescent athletes are significantly less likely to report a concussion than females [51–53]; and the different effects of concussion-induced impulsivity and executive dysfunction on typical male behavioral patterns and aggressive behaviors [54].

Clinical significance and prevention strategies

This meta-analysis reveals a clear association between concussion and an increased risk of suicidal behaviors, which requires a transformation in clinical management models. Concussion should no longer be viewed merely as an acute, self-limiting injury, but should instead be redefined as a complex bio-psycho-social event that can trigger persistent and severe neuropsychiatric sequelae [55]. In clinical practice, it is recommended that all concussion patients receiving medical attention should undergo suicide risk screening at their initial visit, at discharge, and during longitudinal follow-up. Routine, systematic suicide risk screening for concussion patients should become part of the standard clinical pathway [56–58]. Clinicians should maintain a high index of suspicion for suicidality in non-SRC patients. For these individuals, clinical evaluation must go beyond the neurological injury to comprehensively assess the psychosocial context of the injury (e.g., assault, accident), past and current mental health status, and substance use history [59]. Given the significantly elevated suicide risk in males, this group requires particularly close monitoring for post-concussion behavioral changes, expressions of hopelessness, and signs of increased impulsivity. At the same time, the multifaceted nature of post-concussion suicide risk also requires a move away from isolated care towards an integrated, multidisciplinary model that combines neurology, psychiatry, rehabilitation medicine, and social work [60, 61] to comprehensively address the complex needs of patients. From a public health perspective, these findings redefine concussion prevention as a key suicide prevention strategy [62].

Strengths and limitations

The strengths of this study include its large-scale sample size, comprehensive search strategy, strict adherence to PRISMA guidelines, and systematic assessment of publication bias and result robustness. However, this study also has limitations. First, the significant heterogeneity observed in the analysis, particularly in the subgroup analysis of non-sport-related concussion, indicates that the pooled estimates should be interpreted with caution. This variability may stem from differences in study populations, concussion diagnostic criteria, outcome measures, and the degree of adjustment for confounding variables. Second, the study relies on observational studies, mainly of cross-sectional design, which hinders causal inference. Third, many included studies used self-report methods to determine concussion history and suicidal behaviors, which may be affected by recall and reporting bias. Although publication bias was detected in the Suicidal Ideation analysis, the trim-and-fill analysis showed that it did not substantially change the significance of the study’s results, indicating that the result is robust.

Conclusion

This systematic review and meta-analysis of over 650,000 patients with concussion found that both sport-related and non-sport-related concussion are risk factors for suicidal behaviors, including Suicidal Ideation, Made a Suicide Plan, and Attempted Suicide, with comparable risk levels. The findings underscore the necessity of integrating suicide prevention measures into long-term concussion management strategies and call for a collaborative effort among clinicians, researchers, and policymakers to mitigate these preventable outcomes.

Authors’ contributions

MYL conceived and designed the research project. MYL and ELH searched and screened the studies. All authors carried out the risk of bias assessment and data extraction. KYL analysed the data and interpreted the results of the analysis. MYL produced the initial draft of the manuscript. All authors drafted the manuscript and assisted in the interpretation of the data. All authors critically reviewed the manuscript, provided significant input to the initial submission and subsequent revisions, and approved the final version. KYL is responsible for the overall content as the guarantor.

Funding

None.

Data availability

The datasets used and/or analysed during the current study are available fromthe corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.